The Extra-Solar Union of Systems Archive

Otagia wrote:You do realise that the space shuttle moves at 7680 m/s, right?

Arizona Nova wrote:Alright. Kertai, the digisent that helps Anikar, is now going to use the BattleNet to coordinate a devastating attack on the Departed. However, he can only do so if you allow him very temporary control of your ships. He will use information he harvested from the Departed (as described in the post) that highlights all their weaknesses. Skeelzania and Zerstorendar will be the only ones for whom this will not be directly possible, but he will send you both everything you need to know. You all need to react to that request, and you can say no. However, whatever of your force remains will be in the same peril as it was before. Those who say yes will be able to save their remaining forces.

Skeelzania wrote:I figure a few trillion tons of mass that you might have to turn at some point isn't something you want going at supersonic speeds. And 300m/s seems plenty fast to me; lets my Dreadnoughts cover their own length in 4.8 seconds.

Anyway, I'm pretty sure a comprehensive list of the enemy forces hasn't been posted. Its somewhere between "assload" and "gibber-inducing," from what I can tell.

Arizona Nova wrote: Kanuck, the point of Kertai doing this is that he is temporarily acting on the plot-device level with the enemy, and utilizing loopholes in the software and hardware. Your own systems can even start in on him and try to evict him even as he goes for it; it'll make things interesting. Seeing as you can just escape, though, it would make sense if your captain didn't feel the same urgent need to either fight or die, although this is still going to conflict, I would believe, with their horror at seeing them slaughter civilians from orbit.

Kanuckistan, last November wrote: The Kanuckistani approach to the problem - that is, running into someone whose computer tech is advanced enough to bypass your security - is an inversion of the challenge. Basicly, if a system is connected, directly or indirectly, to a potential intrusion vector, then it'll all be hardwired; totally inflexible, only physically capable of performing very specific task tasks, with everything broken down into simple roles. Wanna reprogram? Then you've got to physically rewire the system.

But you need flexibility, if only as a backup, so more traditional systems are connected via devices only capable of passing very specific information(ie, Input X is converted to Y = anything not in line with expectations becomes garbled nonsense) and run in otherwise isolation, with any input run on a virtual machine in isolation from the system it's physically on. We then have another system called "overwatch", which is disconnected from everything equipped with scanners capable of reading the state of it's assigned system - any intruder has no way of getting to the Overwatch system, or hiding from it. Overwatch may in turn have integrated means of dealing with malicious software, or in more sensitive applications, alerts an operator with physical access to the affected hardware - who then yanks it, and sends it down for analysis.

And of course my AIs are build with similar barriers - generally a three tiered system, taking the role of overwatch, with a second sub-sapient Overwatch above them.